agentry/drafter
Drafts a QuickAgent from a description, searching the registry and the web; AI Review reads every result first.
Built into Agentry. The application runs this package itself; it is published here so you can read how it works. Its README says what Agentry gives it beyond its manifest.
Declared capabilities
Ships knowledge: drafting-guide (211.3 KB), in the release archive under knowledge/<name>/, scanned with the version.
Capabilities are the author's declaration and the basis for policy and scanning. The registry cannot enforce them at run time. Your runtime should.
MCP servers & tools
agentry run serves exactly the tools this contract declares, through a per-run MCP gateway. An undeclared capability is a tool that does not exist at run time.
Tools the runner grants
| Manifest declaration | Tool(s) granted |
|---|---|
| (always) | write_report — reports → the run's output folder |
| (always) | list_inputs, read_input, search_inputs — operator-provided input files, read-only |
MCP servers involved
- QuickAgent Tools — the runner's companion server, served through a per-run Docker MCP Gateway; its tool surface is exactly the table above.
In an embedded install
Installed into an existing coding agent (agentry install), the same contract maps to the host's built-in tools — enforcement there is the host's. Allowed tools: none declared (host defaults apply).
Install
agentry install agentry/drafter@1.15.4git clone https://secureagentry.net/agentry/drafter.gitgit clone ssh://git@git.secureagentry.net/agentry/drafter.gitclaude-skill is this agent's default target, so the flag is not needed.
knowledge/drafting-guide/wiki/index.md
--- title: Index type: index summary: Every page in drafting-guide, one line each, by section. Generated; do not edit by hand. --- # drafting-guide 19 pages. ## Draft - [Worked drafts](draft/examples.md) — Five whole drafts, each from a description to the manifest, the shape of its instructions and its notes; one with no tools, one with fetch and a key, one with a declared MCP server, each modelled on one of Agentry's first-party example QuickAgents, one that checks a server the person names (a target host), and one that checks a website the person names in a browser. - [Writing the instructions](draft/instructions.md) — How to write the AGENT.md a draft returns, the instruction its QuickAgent's agent follows; the six headings in order, what each holds, how much to write, and what to leave out because the System Prompt already says it. - [What checks a draft](draft/review.md) — The checks a draft meets after you answer (the rules, AI Review's check of the finished draft, one repair turn) and the scan that decides whether a submitted QuickAgent publishes, goes to review or is rejected. Today's facts; rewritten when the scan's rules on describing risk change. ## Guide - [Start here](guide/start-here.md) — What this guide is, that it is yours and not the person's, and which page to read at each step of a draft. ## Manifest - [A browser](manifest/browser.md) — When a QuickAgent tests a website in a real browser (pages that render, forms that work, links that resolve, a clean console, a phone's width), how it declares one (a url question listed in capabilities.browser.sites, the hosts its pages load from, the scope, the keys it may type), what the run has for it (the browser tools, screenshots in reports/browser/), how to write its instructions, and where it runs. - [Network and fetch](manifest/egress-and-fetch.md) — When a QuickAgent needs the network, how its own fetch tool works (methods, keys, redirects, the response cap, saving a body to the scratch folder), how to choose and name its hosts, and which hosts go on an MCP server instead. - [Commands](manifest/exec.md) — When a QuickAgent declares commands at all (computation the tools cannot do, never reading its inputs), the base commands every run provides, where a command can write, what a run In Agentry provides, and Unsafe Mode on the desktop. - [Input files](manifest/inputs.md) — How the person's files reach a run, a repository or a file at an address that the runner fetches for it, the two views the input tools read them in, the manifest's inputs block that says what a QuickAgent takes and which view it reads first, the cases and what each declares, and how a draft asks for its files. - [The intake and the guide](manifest/intake-and-guide.md) — The questions a QuickAgent asks before each run (types, a repository, a file at an address, a site to test and a target host among them, ids, when to ask at all) and the guide a person reads before a run (overview, prepare, files, task, connect), with their bounds and the fields to leave out. - [The person's knowledge at run time](manifest/knowledge-accepts.md) — When a QuickAgent asks the person to attach a knowledge pack of their own at run time (knowledge.accepts), how to write its prompt, and what goes into the instructions instead. A draft never ships a pack. - [The manifest](manifest/manifest.md) — Every field of the manifest a draft returns, what to write in each, and the rules' bounds by value. Check the draft against the bounds before you return it. - [Run modes](manifest/run-modes.md) — Autonomous and interactive runs, when a draft adds interactive, what a conversation can and cannot do, and the opening line. - [Keys](manifest/secrets.md) — How a QuickAgent declares the keys it needs (its own for fetch, a server's on that server), how to mark each one optional or required, how a run gets them without the model seeing them, and what a run does without an optional one. - [Target hosts](manifest/target-hosts.md) — When a QuickAgent takes a target host (a server the person names when the run starts, to set up, manage or diagnose), how it declares one (a host question listed in capabilities.targetHosts), what the run reaches it with (the host tools and the shell, never a command of its own that makes SSH), how to write its instructions, and where it runs. ## Runtime - [How a QuickAgent run works](runtime/how-a-run-works.md) — What a run of the QuickAgent you draft can and cannot do, what the System Prompt already tells it, and how a run on the person's computer, a run In Agentry and a desktop run in Unsafe Mode differ. ## Servers - [Does not run here](servers/does-not-run-here.md) — Declared MCP servers seen to fail in the runner, and packages whose presence as a direct dependency means a server cannot run, each with the reason, the runtime and the date. Generated from servers.yaml; do not edit by hand. - [Declared MCP servers](servers/mcp-servers.md) — When a server that needs a key or an account is the right one (only when nothing key-free does the task, and never a hosted browser), what a declared MCP server meets in the runner (a package installed, started and reached, or a vendor's remote server reached at its address), what that rules out, how a server signs in, and how to check a server before you pin it. Read it before you declare any server. - [Runs here](servers/runs-here.md) — Declared MCP servers seen to install, start and list their tools in the runner, each with the versions, the runtime, the date and what a QuickAgent must declare for it. Generated from servers.yaml; do not edit by hand. - [Remote servers and sign-ins](servers/sign-ins.md) — Vendors' own remote MCP servers, by address and how each signs in, and the package servers' OAuth sign-ins seen working, which a draft copies exactly. Generated from servers.yaml; do not edit by hand.